The proposal presents a general theory of lexical information processing during sentence comprehension which accounts for the processing of ambiguous and unambiguous lexical items with the same set of principles. The theory, The Minimal Information Hypothesis, also makes predictions about the effects of context occurring both before and after lexical items in discourse. Six experiments are proposed to test predictions from the theory, as well as some tentative pilot projects.